Search

We demonstrate low energy, forming and compliance-free operation of a resistive memory obtained by the partial oxidation of a two-dimensional layered van-der-Waals semiconductor: hafnium disulfide (HfS$_2$). Semiconductor - oxide heterostructures are achieved by low temperature ($<300^{o}$C) thermal oxidation of HfS$_2$ in dry conditions, carefully controlling process parameters. The resulting HfO$_x$S$_y$/HfS$_2$ heterostructures are integrated between metal contacts, forming vertical crossbar devices. Forming-free, compliance-free resistive switching between non-volatile states is demonstrated by applying voltage pulses and measuring the current response in time. We show non-volatile memory operation with an R$_{ON}$/ R$_{OFF}$ of 102, programmable by 80ns WRITE and ERASE operations. Multiple stable resistance states are achieved by modulating pulse width and amplitude, down to 60ns, $<$ 20pJ operation. This demonstrates the capability of these devices for low - energy, fast-switching and multi-state programming. Resistance states were retained without fail at 150$^o$C over 10$^4$s, showcasing the potential of these devices for long retention times and resilience to ageing. Low-energy resistive switching measurements were repeated in vacuum (8.6 mbar) showing unchanged characteristics and no dependence of the device on surrounding oxygen or water vapour. Using a technology computer-aided design (TCAD) tool, we explore the role of the semiconductor layer in tuning the device conductance and driving gradual resistive switching in 2D HfO$_x$ - based devices.

Discerning between authentic content and that generated by advanced AI methods has become increasingly challenging. While previous research primarily addresses the detection of fake faces, the identification of generated natural images has only recently surfaced. This prompted the recent exploration of solutions that employ foundation vision-and-language models, like CLIP. However, the CLIP embedding space is optimized for global image-to-text alignment and is not inherently designed for deepfake detection, neglecting the potential benefits of tailored training and local image features. In this study, we propose CoDE (Contrastive Deepfake Embeddings), a novel embedding space specifically designed for deepfake detection. CoDE is trained via contrastive learning by additionally enforcing global-local similarities. To sustain the training of our model, we generate a comprehensive dataset that focuses on images generated by diffusion models and encompasses a collection of 9.2 million images produced by using four different generators. Experimental results demonstrate that CoDE achieves state-of-the-art accuracy on the newly collected dataset, while also showing excellent generalization capabilities to unseen image generators. Our source code, trained models, and collected dataset are publicly available at: https://github.com/aimagelab/CoDE., Comment: ECCV 2024

Thermoelectric materials are of great interest for heat energy harvesting applications. One such promising material is TlGaSe$_{2}$, a p-type semiconducting ternary chalcogenide. Recent reports show it can be processed as a thin film, opening the door for large-scale commercialization. However, TlGaSe$_{2}$ is prone to stacking faults along the [001] stacking direction and their role in its thermoelectric properties has not been understood to date. Herein, TlGaSe$_{2}$ is investigated via (scanning) transmission electron microscopy and first-principles calculations. Stacking faults are found to be present throughout the material, as density functional theory calculations reveal a lack of preferential stacking order. Electron transport calculations show an enhancement of thermoelectric power factors when stacking faults are present. This implies the presence of stacking faults is key to the material's excellent thermoelectric properties along the [001] stacking direction, which can be further enhanced by doping the material to hole carrier concentrations to approx. 10$^{19}$ cm$^{-3}$.

The objective of image captioning models is to bridge the gap between the visual and linguistic modalities by generating natural language descriptions that accurately reflect the content of input images. In recent years, researchers have leveraged deep learning-based models and made advances in the extraction of visual features and the design of multimodal connections to tackle this task. This work presents a novel approach towards developing image captioning models that utilize an external kNN memory to improve the generation process. Specifically, we propose two model variants that incorporate a knowledge retriever component that is based on visual similarities, a differentiable encoder to represent input images, and a kNN-augmented language model to predict tokens based on contextual cues and text retrieved from the external memory. We experimentally validate our approach on COCO and nocaps datasets and demonstrate that incorporating an explicit external memory can significantly enhance the quality of captions, especially with a larger retrieval corpus. This work provides valuable insights into retrieval-augmented captioning models and opens up new avenues for improving image captioning at a larger scale., Comment: ACM Transactions on Multimedia Computing, Communications and Applications (2024)

Oxidation of two-dimensional (2D) layered materials has proven advantageous in creating oxide/2D material heterostructures, opening the door for a new paradigm of low-power electronic devices. Gallium (II) sulfide ($\beta$-GaS), a hexagonal phase group III monochalcogenide, is a wide bandgap semiconductor with a bandgap exceeding 3 eV in single and few layer form. Its oxide, gallium oxide (Ga$_2$O$_3$), combines large bandgap (4.4-5.3 eV) with high dielectric constant (~10). Despite the technological potential of both materials, controlled oxidation of atomically-thin $\beta$-GaS remains under-explored. This study focuses into the controlled oxidation of $\beta$-GaS using oxygen plasma treatment, achieving ultrathin native oxide (GaS$_x$O$_y$, ~4 nm) and GaS$_x$O$_y$/GaS heterostructures where the GaS layer beneath remains intact. By integrating such structures between metal electrodes and applying electric stresses as voltage ramps or pulses, we investigate their use for resistive random-access memory (ReRAM). The ultrathin nature of the produced oxide enables low operation power with energy use as low as 0.22 nJ per operation while maintaining endurance and retention of 350 cycles and 10$^4$ s, respectively. These results show the significant potential of the oxidation-based GaS$_x$O$_y$/GaS heterostructure for electronic applications and, in particular, low-power memory devices.

In recent years, remarkable advancements have been achieved in the field of image generation, primarily driven by the escalating demand for high-quality outcomes across various image generation subtasks, such as inpainting, denoising, and super resolution. A major effort is devoted to exploring the application of super-resolution techniques to enhance the quality of low-resolution images. In this context, our method explores in depth the problem of ship image super resolution, which is crucial for coastal and port surveillance. We investigate the opportunity given by the growing interest in text-to-image diffusion models, taking advantage of the prior knowledge that such foundation models have already learned. In particular, we present a diffusion-model-based architecture that leverages text conditioning during training while being class-aware, to best preserve the crucial details of the ships during the generation of the super-resoluted image. Since the specificity of this task and the scarcity availability of off-the-shelf data, we also introduce a large labeled ship dataset scraped from online ship images, mostly from ShipSpotting\footnote{\url{www.shipspotting.com}} website. Our method achieves more robust results than other deep learning models previously employed for super resolution, as proven by the multiple experiments performed. Moreover, we investigate how this model can benefit downstream tasks, such as classification and object detection, thus emphasizing practical implementation in a real-world scenario. Experimental results show flexibility, reliability, and impressive performance of the proposed framework over state-of-the-art methods for different tasks. The code is available at: https://github.com/LuigiSigillo/ShipinSight ., Comment: Accepted at 2024 International Joint Conference on Neural Networks (IJCNN)

The recent surge in 3D data acquisition has spurred the development of geometric deep learning models for point cloud processing, boosted by the remarkable success of transformers in natural language processing. While point cloud transformers (PTs) have achieved impressive results recently, their quadratic scaling with respect to the point cloud size poses a significant scalability challenge for real-world applications. To address this issue, we propose the Adaptive Point Cloud Transformer (AdaPT), a standard PT model augmented by an adaptive token selection mechanism. AdaPT dynamically reduces the number of tokens during inference, enabling efficient processing of large point clouds. Furthermore, we introduce a budget mechanism to flexibly adjust the computational cost of the model at inference time without the need for retraining or fine-tuning separate models. Our extensive experimental evaluation on point cloud classification tasks demonstrates that AdaPT significantly reduces computational complexity while maintaining competitive accuracy compared to standard PTs. The code for AdaPT is made publicly available., Comment: 26 pages, 8 figures, submitted to Neural Networs

The recent surge of pervasive devices that generate dynamic data streams has underscored the necessity for learning systems to adapt continually to data distributional shifts. To tackle this challenge, the research community has put forth a spectrum of methodologies, including the demanding pursuit of class-incremental learning without replay data. In this study, we present MIND, a parameter isolation method that aims to significantly enhance the performance of replay-free solutions and achieve state-of-the-art results on several widely studied datasets. Our approach introduces two main contributions: two alternative distillation procedures that significantly improve the efficiency of MIND increasing the accumulated knowledge of each sub-network, and the optimization of the BachNorm layers across tasks inside the sub-networks. Overall, MIND outperforms all the state-of-the-art methods for rehearsal-free Class-Incremental learning (with an increment in classification accuracy of approx. +6% on CIFAR-100/10 and +10% on TinyImageNet/10) reaching up to approx. +40% accuracy in Domain-Incremental scenarios. Moreover, we ablated each contribution to demonstrate its impact on performance improvement. Our results showcase the superior performance of MIND indicating its potential for addressing the challenges posed by Class-incremental and Domain-Incremental learning in resource-constrained environments., Comment: Accepted at the 38th AAAI Conference on Artificial Intelligence

Machine Unlearning is rising as a new field, driven by the pressing necessity of ensuring privacy in modern artificial intelligence models. This technique primarily aims to eradicate any residual influence of a specific subset of data from the knowledge acquired by a neural model during its training. This work introduces a novel unlearning algorithm, denoted as Distance-based Unlearning via Centroid Kinematics (DUCK), which employs metric learning to guide the removal of samples matching the nearest incorrect centroid in the embedding space. Evaluation of the algorithm's performance is conducted across various benchmark datasets in two distinct scenarios, class removal, and homogeneous sampling removal, obtaining state-of-the-art performance. We also introduce a novel metric, called Adaptive Unlearning Score (AUS), encompassing not only the efficacy of the unlearning process in forgetting target data but also quantifying the performance loss relative to the original model. Additionally, we conducted a thorough investigation of the unlearning mechanism in DUCK, examining its impact on the organization of the feature space and employing explainable AI techniques for deeper insights.

This work considers the problem of approximating initial condition and time-dependent optimal control and trajectory surfaces using multivariable Fourier series. A modified Augmented Lagrangian algorithm for translating the optimal control problem into an unconstrained optimization one is proposed and two problems are solved: a quadratic control problem in the context of Newtonian mechanics, and a control problem arising from an odd-circulant game ruled by the replicator dynamics. Various computational results are presented. Use of automatic differentiation is explored to circumvent the elaborated gradient computation in the first-order optimization procedure. Furthermore, mean square error bounds are derived for the case of one and two-dimensional Fourier series approximations, inducing a general bound for problems of $n$ dimensions.

Semantic representation is a key enabler for several application domains, and the multi-agent systems realm makes no exception. Among the methods for semantically representing agents, one has been essentially achieved by taking a behaviouristic vision, through which one can describe how they operate and engage with their peers. The approach essentially aims at defining the operational capabilities of agents through the mental states related with the achievement of tasks. The OASIS ontology -- An Ontology for Agent, Systems, and Integration of Services, presented in 2019 -- pursues the behaviouristic approach to deliver a semantic representation system and a communication protocol for agents and their commitments. This paper reports on the main modeling choices concerning the representation of agents in OASIS 2, the latest major upgrade of OASIS, and the achievement reached by the ontology since it was first introduced, in particular in the context of ontologies for blockchains., Comment: Please cite https://www.scopus.com/record/display.uri?eid=2-s2.0-85165473819&origin=resultslist

The Bass diffusion equation is a well-known and established modeling approach for describing new product adoption in a competitive market. This model also describes diffusion phenomena in various contexts: infectious disease spread modeling and estimation, rumor spread on social networks, prediction of renewable energy technology markets, among others. Most of these models, however, consider a deterministic trajectory of the associated state variable (e.g., market-share). In reality, the diffusion process is subject to noise, and a stochastic component must be added to the state dynamics. The stochastic Bass model has also been studied in many areas, such as energy markets and marketing. Exploring the stochastic version of the Bass diffusion model, we propose in this work an approximation of (stochastic) optimal control surfaces for a continuous-time problem arising from a $2\times2$ skew symmetric evolutionary game, providing the stochastic counter-part of the Fourier-based optimal control approximation already existent in the literature.

Nickel-iron layered double hydroxide (NiFe LDH) platelets with high morphological regularity and sub-micrometre lateral dimensions were synthesized using a homogeneous precipitation technique for highly efficient catalysis of the oxygen evolution reaction (OER). Considering edge sites are the point of activity, efforts were made to control platelet size within the synthesized dispersions. The goal is to controllably isolate and characterize size-reduced NiFe LDH particles. Synthetic approaches for size control of NiFe LDH platelets have not been transferable based on published work with other LDH materials and for that reason, we instead use post-synthetic treatment techniques to improve edge-site density. In the end, size reduced NiFe LDH/single-wall carbon nanotube (SWCNT) composites allowed to further reduce the OER overpotential to 237 plus/minus 7 mV ( L = 0.16 plus/minus 0.01 micrometres, 20 wt% SWCNT), which is one of the best values reported to date. This approach as well improved the long term activity of the catalyst in operating conditions.

MXenes are a class of 2D/layered materials which are highly conductive, hydrophilic, have a large electrochemical surface area and are easily processible into electrodes for energy applications. Since the discovery of MXenes over ten years ago, these materials have been mainly used in the preparation of electrodes for batteries and supercapacitors. However, due to their aforementioned properties, MXenes could potentially be utilised as a component in the catalyst layer for the Oxygen Evolution Reaction (OER). This opinion piece will discuss some of the recent literature in the area of hybrid catalysts consisting of various Transition Metal Oxides (TMOs) and MXenes for the OER. We will also discuss current drawbacks and future outlook in this new area of research.